hive-cli
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill provides instructions for using
npxto fetch and execute thehive-intelligencepackage directly from the npm registry to initialize the environment.\n- [EXTERNAL_DOWNLOADS]: Recommends installing vendor packages from npm and provides links to official documentation and tutorials hosted on thehiveintelligence.xyzdomain.\n- [PERSISTENCE]: Includes instructions for scheduling shell scripts viacronto perform daily data retrieval tasks.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from Hive API endpoints (ingestion points inSKILL.mdCLI queries) and interpolates it into generated shell scripts (capabilities include bash execution) without explicit boundary markers or sanitization, creating a surface for indirect prompt injection.
Audit Metadata